We are attempting to improve our previously published theoretical models of the effect of macromolecular crowding upon protein associations and isomerizations. Current work focuses on the statistical-thermodynamic formulation of a potential of average force acting between two "tracer" hard particles in a fluid containing an arbitary volume fraction of "crowder" hard particles. This is not an analytically soluble problem, and we are trying a variety of approximate techniques in order to arrive at a robust solution that is not critically dependent upon the nature of the approximations made in the formulation of the model. We plan to compare the results of the theoretical models with the results of Monte Carlo and/or molecular dynamics simulation.